二维小波变换去噪的MATLAB可视化设计

MATLAB Visual Design for Denoising via 2D Wavelet Transform

小波变换是消除图像和多道信号噪声的一种重要方法,而MATLAB是优秀的数值计算和可视化设计平台。二维小波变换将多道记录信号变换到时间、频率、空间和波数四维域中,根据信号和相干噪声在频率、时间、波数、空间等方面的局部差异来达到滤波的目的。基于二维多分辨分析的基本原理,利用MATLAB的图形用户界面(GUI)和丰富的图像处理功能进行可视化设计,实现了用二维小波变换对多道信号中相干噪声的消除,使烦琐的逐级分解去噪处理变得容易,提高了工作效率。最后,用设计的程序对地震记录信号中的面波进行了消除,并达到了较好的效果。

Wavelet transform is an important denoising method and MATLAB is an excellent tool of computing and visual design. 2D wavelet transform change multi channel signal into 4D domain which is consists of time, frequency, space and wave number. On the basis of local difference between signal and noise, filtering was done. In the light of 2D multi resolution theory, by means of MATLAB, this paper eliminated coherent noise via 2D wavelet transform. As an example, test on ground roll elimination was done. The result showed that the process was convenient and its efficient was high.